[Eeglablist] Component scalp maps and dipoles

[Robert Oostenveld]

On 28 Sep 2006, at 18:38, Victoria Eugenia Montes Restrepo wrote:
> I’m using the DIPFIT plug-in to obtain dipoles from the independent  
> components given by ICA. I want to know what is the physical  
> meaning of the component scalp maps and what is their relation with  
> the dipoles, I mean how are the dipoles found from these maps?

Dear Victoria,

You should look at the first section of http://www.sccn.ucsd.edu/ 
eeglab/dipfittut/dipfit.html if you have not done that yet. I find  
the 1990 paper from Scherg (creator of BESA, see reference below)  
good as a general introduction. Regarding that paper: it is described  
there how to make complex dipole models to explain all spatio- 
temporal features in the data. The approach taken in EEGLAB is to use  
ICA to disentangle the spatial source components from the temporal  
source components. That makes the dipole modelling much easier, since  
the dipoles only have to explain the spatial distribution of each  
component seperately, and not make a complex many-dipole model for  
the complete data as can be done in BESA. The timcourse of the dipole  
in EEGLAB simply corresponds to the timecourse of the ICA component  
to which the dipole was fitted.

best regards,
Robert

---
M. Scherg, "Fundamentals of dipole source potential analysis". In:  
"Auditory evoked magnetic fields and electric potentials" (eds. F.  
Grandori, M. Hoke and G.L. Romani). Advances in Audiology, Vol. 6.  
Karger, Basel, pp 40-69, 1990.